Teaching with FCC v1.5¶
The FCC v1.5.x line introduced six architectural pillars, a codename decoder, an expanded 164-persona catalog, and a full SAFe operating model. Any of those alone is a semester's worth of material. The challenge for an educator is not finding enough to teach but deciding what to leave out. This page offers one concrete answer: a 15-week beginner arc, an assessment rubric drawn from the PHOENIX 58-gate quality registry, a mapping from student-cohort archetypes onto FCC personas, and a lab exercise bank aligned with the curriculum.
Use this page alongside the existing 12-Week University Curriculum and the Lab Exercise Bank. The 12-week curriculum remains the default; this 15-week arc is the beginner-focused alternative that a first-semester course can adopt when the class mix is heavier on CS-first-year than on senior-level capstone.
Course philosophy — three teaching claims¶
Before adopting this arc, decide whether these three claims match your program's pedagogical stance.
- Agentic AI is a team metaphor, not a magic metaphor. The easiest way for a student to understand FCC is to think of it as a project team with strict roles and a well-documented communication protocol. The "AI" part is secondary.
- Persona-first is more scaffolding than concept-first. Students new to agentic systems benefit from handling specific personas early (Research Crafter, Blueprint Crafter, Governance Compliance Auditor) before they meet the abstract R.I.S.C.E.A.R. contract.
- Reproducibility is a teachable, testable discipline. Every lab in this arc is deterministic; every student submission should reproduce bit-identically on the instructor's machine.
If those three align with your program, the rest of this page is prescriptive. If not, pick and choose.
The 15-week beginner arc¶
The arc assumes 2 hours of lecture + 2 hours of lab per week, 15 weeks total. It is designed for a 3- or 4-credit course.
Weeks 1–3: Orientation¶
| Week | Lecture topic | Lab |
|---|---|---|
| 1 | What is FCC? Install, verify, hello-fcc | first-30-minutes.md walkthrough |
| 2 | Personas as structured roles | Notebook 03_persona_exploration.ipynb |
| 3 | The Find → Create → Critique cycle | Notebook 02_workflow_basics.ipynb |
Learning outcomes by Week 3: student can install FCC, list the categories, run a mock simulation, and describe the FCC cycle in plain language.
Weeks 4–6: Personas in depth¶
| Week | Lecture topic | Lab |
|---|---|---|
| 4 | The R.I.S.C.E.A.R. spec (10 elements) | Author a minimal custom persona |
| 5 | Discernment + design-target factors | Dimension profile for a custom persona |
| 6 | Champions and orchestration | Notebook 12_custom_persona_design.ipynb |
Learning outcomes by Week 6: student can author a well-formed
R.I.S.C.E.A.R. YAML that passes fcc audit personas --strict, justify
each of the 10 elements, and wire a champion persona that orchestrates
2-3 base personas.
Weeks 7–9: The v1.5 pillars (light touch)¶
| Week | Lecture topic | Lab |
|---|---|---|
| 7 | Pillar A + E: Seed data + Performance | Seed-dry-run + benchmark runner |
| 8 | Pillar B: GraphRAG | Notebook 46_graphrag_advanced_patterns.ipynb |
| 9 | Pillar D: CRDT multi-user | Your First Multi-User Session |
Learning outcomes by Week 9: student can describe each pillar at a one-paragraph level, run a GraphRAG query, and observe a CRDT merge between two simulated users.
Weeks 10–12: Governance and compliance¶
| Week | Lecture topic | Lab |
|---|---|---|
| 10 | Quality gates + tiered findings | fcc audit personas --strict on a custom persona |
| 11 | EU AI Act + NIST AI RMF | Classify a custom persona with AIActClassifier |
| 12 | Model cards + transparency cards | Generate cards for a custom persona |
Learning outcomes by Week 12: student can explain why governance is a first-class FCC concern, read a model card, and produce a minimal compliance report for a custom persona.
Weeks 13–15: Capstone¶
| Week | Activity |
|---|---|
| 13 | Capstone design review: each student presents a 3-persona team + scenario |
| 14 | Capstone build: students implement their scenario end-to-end |
| 15 | Capstone presentation + peer critique using the FCC Critique phase |
Capstone grading follows the rubric in the next section.
Assessment rubrics using PHOENIX quality gates¶
The PHOENIX 58-gate extended registry (vendored in v1.3.8) is
already a natural assessment rubric. The tier labels — tier_1,
tier_2, tier_3 — map directly to "Must Fix", "Should Fix", and
"Nice to Have". Use them to grade custom-persona submissions.
Rubric template for a custom persona assignment¶
| Gate | Tier | Points | Expected evidence |
|---|---|---|---|
| QG-RC-001 | 1 | 20 | Persona YAML passes fcc audit personas --strict |
| QG-DGS-001 | 2 | 10 | data_flow frontmatter is complete |
| QG-AMS-001 | 1 | 10 | No hallucinated collaborator IDs |
| Dimensions profile present | 2 | 10 | 56-dimension profile matches the schema |
| Model card generated | 2 | 10 | ModelCardGenerator.from_persona succeeds |
| EU AI Act classification | 2 | 10 | AIActClassifier assigns a category |
| At least one action references persona | 3 | 10 | Action engine trace succeeds |
| Adoption checklist is concrete | 3 | 10 | Checklist items are testable |
| Discernment matrix complete | 3 | 10 | All 6 traits × 3 ratings present |
| Total | 100 |
This rubric is automation-friendly. A grading harness can run
fcc audit personas --strict and collect the findings; partial credit
is awarded based on the tier + severity of remaining findings.
Rubric for the capstone¶
A capstone submission is a 3-persona team plus a scenario plus a run trace. The rubric extends the per-persona grading with scenario-level gates:
| Dimension | Weight | What is graded |
|---|---|---|
| Persona quality (×3) | 45% | Each persona graded with the template above; average |
| Scenario completeness | 15% | Scenario JSON validates; seed data is declared |
| Run trace reproducibility | 15% | Re-run produces bit-identical events |
| Critique quality | 15% | Student's own critique identifies at least three plausible improvements |
| Documentation | 10% | README describes the team + scenario + run instructions |
Student cohorts mapped to FCC personas¶
Students don't all come to class from the same background. Mapping cohort archetypes onto FCC personas helps students find an in-framework role that matches their strengths.
| Cohort archetype | FCC persona affinity |
|---|---|
| The implementer (loves code) | Developer personas — Implementation Engineer, Plugin Developer |
| The writer (loves clarity) | Documentation Strategist, Three-Layer Documentation Reconciler |
| The systems thinker (loves architecture) | Blueprint Crafter, Integration Architect |
| The auditor-in-training (loves rules) | Governance Compliance Auditor, Forensic Auditor, Patent Claim Tracer |
| The researcher (loves ideas) | Research Crafter, Research Handoff Liaison |
| The planner (loves schedules) | SAFe Metrics Crafter, Long-Term Archive Steward |
| The teacher (loves explanation) | Knowledge Graph Serializer Elevator, Decision Archaeologist |
Use this mapping for team formation in the capstone. A well-balanced capstone team includes at least three distinct cohort archetypes.
Lab exercises aligned with the 15-week arc¶
The existing Lab Exercise Bank ships 15 exercises across beginner, intermediate, and advanced tiers. For the 15-week arc, here's the recommended weekly pairing.
| Week | Lab from bank |
|---|---|
| 1 | B1 — Inventory the persona catalog |
| 2 | B2 — Cross-reference a persona |
| 3 | B3 — Run a mock simulation |
| 4 | B4 — Author a minimal persona |
| 5 | B5 — Add a dimensions profile |
| 6 | I1 — Author a champion persona |
| 7 | I2 — Run a benchmark |
| 8 | I3 — GraphRAG query (mapped to notebook 46) |
| 9 | I4 — Observe a CRDT merge (mapped to the beginner CRDT walk-through) |
| 10 | I5 — Strict audit on a custom persona |
| 11 | A1 — EU AI Act classification |
| 12 | A2 — Generate and read a model card |
| 13 | A3 — Capstone design review |
| 14 | A4 — Capstone build |
| 15 | A5 — Capstone critique |
Swap in lab exercises from the bank as needed; the mapping above is prescriptive but not rigid.
Workshop variants for professional audiences¶
The 15-week arc condenses cleanly into four shorter formats for professional training.
2-day intensive workshop¶
Drawn from Workshop 2-Day. Day 1 covers Weeks 1-6 (orientation + personas in depth). Day 2 covers Weeks 7-12 (pillars + governance). Capstone is optional homework.
1-day executive briefing¶
Weeks 1, 6, 9, 12 as a tour. The focus is on the codename decoder,
the persona catalog, the v1.5.0 pillars, and the governance
surface. No hands-on labs; use screenshots and notebook 39
(v14x_guided_tour) as a live demo.
5-day bootcamp¶
Every day covers 3 weeks of the 15-week arc. Compressed, intense; requires prior Python experience. Capstone is the final afternoon of Day 5.
1-semester graduate seminar¶
Replace the beginner-focused Weeks 1-3 with an intensive first week of "Read the FCC Guidebook chapters 1-5 in full and present." Use the remaining 14 weeks to cover the 15-week arc Weeks 4-15.
Using the codename decoder in the classroom¶
The Codename Decoder (v1.5.1) is an unexpectedly rich teaching asset. Each codename carries its mythological or astronomical origin story, which gives students a memorable hook. Some teaching moves:
- Week 1 icebreaker. Ask each student to pick one codename from the decoder and present the origin story to the class. By Monday of Week 2, the class has shared ownership of the ecosystem vocabulary.
- Week 7 Pillar B anchor. When introducing GraphRAG, lead with LYRA — Orpheus's lyre — and the harmonic-expansion metaphor. Students remember "harmonic expansion" better than "SKOS-proxy edges".
- Week 10 governance anchor. PHOENIX — the self-regenerating firebird — is the perfect metaphor for a continuously-rebuilt quality registry. Governance feels less arbitrary with this framing.
- Capstone naming. Assign each capstone team a constellation vertical to study (or let them pick). The mythological framing makes the project memorable.
Integration with the guidebook and the book series¶
FCC ships a 24-chapter Guidebook (docs/guidebook/) and a three-book
series (docs/books/). Recommended reading alongside the 15-week arc:
| Arc weeks | Guidebook chapters | Book chapters |
|---|---|---|
| 1-3 | 1, 3 | Understanding FCC ch. 1-3 |
| 4-6 | 3, 4, 26 | Understanding FCC ch. 4-6 |
| 7-9 | 32, 35 | Building with FCC ch. 1-3 |
| 10-12 | 9, 20, 30 | Building with FCC ch. 6-8 |
| 13-15 | 11, 12, 25 | Advanced FCC ch. 1-3 |
The Guidebook chapters 31-35 cover the v1.4.x additions specifically; chapters 36+ (as they ship in future releases) will extend coverage to the six v1.5.0 pillars.
Assessment strategies for non-CS majors¶
Not every class that benefits from FCC is a computer-science class. For students from public policy, biology, law, or business administration, the lab exercises in the bank are too code-heavy. Alternative assessments:
- Policy case study. Students analyse a persona YAML and write a 1,500-word policy memo on how it could be adopted in a regulated sector.
- Ethics presentation. Students pick a persona with a
tier_1EU AI Act classification and present the ethical considerations in a 15-minute talk. - Workflow mapping. Students map an existing workflow in their discipline (e.g., a clinical-trial review process) onto the FCC cycle (Find → Create → Critique) and identify which personas could be useful.
- Codename origin project. Students pick one codename and write a 2,000-word essay on whether the mythological framing is well-chosen (research-ethics bonus: critique the framing).
All four alternatives bypass hands-on coding while retaining the conceptual depth.
Common teaching challenges¶
Students conflate personas with tools¶
The easiest fix: repeatedly emphasise that a persona is a structured role specification, not a piece of code. A persona YAML is to a team member what a job description is to an employee.
Students skip the R.I.S.C.E.A.R. audit¶
The audit is the discipline that makes FCC legible. Make every graded submission require a passing audit; the rubric in this page makes this mechanical.
Students under-invest in the Critique phase¶
Many students treat Find → Create as the "real work" and Critique as an afterthought. Rebalance by making the capstone critique worth 15% of the final grade.
Codenames feel arbitrary at first¶
Use the Codename Decoder from Week 1. Once students internalise the mythology, the names become mnemonic aids rather than barriers.
Related resources¶
- 12-Week University Curriculum — the default curriculum.
- Workshop 2-Day — intensive workshop format.
- Assessment Strategies — grading strategies beyond the rubric on this page.
- Student Workbook — hands-on exercises.
- Lab Exercise Bank — 15 labs mapped to the 15-week arc above.
- Grading Rubrics — companion rubric detail.
Next steps¶
- Review the 12-Week Curriculum and decide which arc fits your class mix.
- Read the Codename Decoder and pick five codenames to lead with in Week 1.
- Walk
docs/tutorials/migration/restructuring-existing-documentation.md(Pool A) for the discipline of restructuring existing material around FCC — useful if you're adopting FCC mid-semester. - Review ADR-010 — Six pillars architecture for the Week 7-9 content anchor.
- Skim
docs/tutorials/sample-prompts/educators-teach-safe-rollout.mdfor existing prompts; swap in the v1.5.x equivalents as they land. - Open notebook
39_v14x_guided_tour.ipynbas the Week 1 lecture companion. - Open notebook
45_v14x_capstone_scenarios.ipynbas the Week 14 lecture companion.